CA1332325C - Stable membranes from sulfonated polyarylethers - Google Patents

Abstract

-i-PATENT APPLICATION OF

John Tomaschke, San Diego, CA
Anthony Joseph Testa, Westwood, MA
James George Vouros, Boston, MA
for STABLE MEMBRANES FROM SULFONATED POLYARYLETHERS
Abstract of the Disclosure The invention provides a novel thin film composite or coated membrane suitable for reverse osmosis, ultrafil-tration and microfiltration applications, and having a porous polymeric substrate with one or more microporous layers to which a thin film or coating comprising a sulfonated polyarylether is attached substantively to provide an oxidatively stable, thin hydrophilic film or coating layer, and a method for manufacturing and using the same.

Description

:

~3~232~

BACKGROUND OF THE INVENTION

FIELD OF THE INVENTION. The present invention relates to composite or coated sulfonated poly(arylether) membranes ~ -~
for reverse osmosis ("RO" ), ultrafiltration ("UF") and mic ofiltration ("MF") made by a novel process and their method of use.
'- :, DESCRIPTION OF T~E PRIOR ART. The present comme-cially available membranes fo~ water desalination by reve_se osmosis a~e derived f om two basic classes of polymers:
cellulosic (predominantly cellulose acetates) and the newe_ generation condensation products consisting of poly~
amides, polyamidohydrazides, polyureas and polyetheru-eas.
The cellulosic membranes are susceptible to mic-obiologi~
cal attac~, compaction at higher temperatures and pres-sures, and are limited to a relatively na~row feed pH
~; range. These membranes do, however, possess fairly good resistance to low levels of chlorine, a popular disinfec-tant and cleaning chemical used in water desalination and other separation processes. The second group of polymers ~20 (which are suitable for ultrafiltration as well as RO) in }`''' ', . . ~',; .. '.. ,, ~,, ':

-2- ~ 33232~

general exhibit improved transport properties at given applied pressures and stability over a wider range of pH
compared to the cellulosic membranes. Unfortunately, all of these membranes made from the newer generation of poly-mers suffer from poor resistance to continual exposure to oxidizing agents such as chlorine in a RO application.
This "tradeoff~ in sensitivities results in economically disadvantageous chlorine removal in many common feed streams or loss of membrane permselectivity due to oxida-tive degradation. Thus, this kind of sensitivity limits or even prevents their use in potable water applications and especially food and beverage, medical, biochemical, and pharmaceutical applications where chlorination and other similar oxidative cleaners or sterilants a~e common-ly employed.
~`
In recent years polysulfone type polymers have been `~ extensively used in the manufacture of ultrafiltration "~' membranes for use in many applications because of excel-lent hydrolytic stability and high temperature properties.
~20 The polysulfone polymers that are commercially available today, UDEL~ polysulfone manufactued by Union Carbide Corp. and Victrex~ polyethersulfone manufactured by ICI, also possess acceptable chlorine resistance when exposed 'periodically during cleaning. The efore, these polysul-~25 fone type polymers are extensively used fo_ memb~ane ap-plications in the dai-y and food processing areas that require daily sanitizations with chlorine and high temper-ature membrane cleaning regimens of 1% caustic and 1%
phosphoric acid. Polysulfone type membranes have also found extensive use in pharmaceutical and biotechnology ; application areas and perform very well under most circum-stances.
.`

:.:

j~, .. ,.. , ~ , , ~ , `~

) 3 ~ 3 3 2 3 2 ~

In recent years, sulfonated polyarylethers, and particularly sulfonated polyarylether sulfones have been examined for membrane separation applications due to their outstanding chemical and thermal stability. In addition to demonstrating good RO membrane flux and salt rejections at pressures greater than 300 psig, these membranes have demonstrated the ability to operate at continuous chlorine exposure levels and pH extremes that would destroy both of the previously mentioned classes of polymer membranes.
Thus the chlorine tolerance and hydrolytic stability of the sulfonated polyarylethers would, at first blush, appear to make them particularly well suited for desalination of a wide range of aggressive feed streams represented by such conditions ranging from natural brackish waters, industSial effluents, sewage effluents, mining waters, agricultural run-off, etc. and many ;~
applications other than water desalination and water recovery mentioned previously.

The problem with these membranes has been thei- ina-2~ bility to achieve commercially attractive fluxes and salt rejections reproducibly at economical applied pressures.
Acceptable performance for brackish water (low p~essure RO) desalination would typically be: at least 15 GFD (gal-lons/ft2-day) water flux and 95% or bette- salt rejection (5% or less salt passage) at 200 psig net driving pressure ;~
(NDP) or alternatively, an equitable tradeoff in these properties. Consistent with the principles of reverse osmosis, higher applied pressures will give higher flux and salt rejection, but at a penalty of added power cost ~30~ for the extra pressure. It is therefore desirable to develop membranes which produce adequate flux and salt ~ ~ rejections at lower applied pressures.
I ~
., ,~ ~

~, ~

i -4- 133~32~

The prior art involving sulfonated polyarylether membranes reveals that in order to obtain the desired transport properties on brackish (2000-5000 ppm NaCl) feeds of 15 or more GFD flux and 95% or better salt rejec-tion, applied pressures of 300-600 psig or even more had to be used. The predominant membrane types developed were asymmet-ic, and consisted of thick ~1-20 mil) anisotropic structures in which the permselectivity occurs at the thin dense film top side of this st-ucture. The thin top film is integral with and supported by a progressively more porous understructure. Better results were found with thin film composite membranes in which selected laboratory samples developed by Graefe, et al (Office of Water Research and Technology, Report Wo. 2001-20) achieved low pressure goal performance at 250 psig (32 GFD flux and 94.5% salt rejection). This latter membrane example consisted of a thin film laminate prepared by brush coating a solution of sulfonated polysulfone onto a porous polysulfone substrate which was pret_eated with aqueous ~23 lactic acid. The purpose for the latte- solution was claimed to prevent intrusion of the sulfonated polysulfone solution into the pores of the porous substrate. This ~ thin film composite membrane while demonst-ating !, ~ potential, suffered from an awkward porous substrate pre-t_eatment step and lack of pe~formance reproductibility ~ and never reached commercialization.
;"
One such thin film composite membrane which was car-ried further in its development was the hollow fiber com-posite system developed by Schiffer, et al~ This membrane ~30 was prepa-ed by coating an alcoholic solution of a highly sulfonated polysulfone (free acid form) onto a porous hollow-fiber polysulfone substrate. The thin film was .

~: ' ~, , ~j ~

claimed to have been crosslinked via the thermal treatment applied after solution deposition, though no proof of this was actually given. Performance was modest - at 6 GFD
flux and up to 95% salt rejection on a 3500 ppm NaCl feed at 400 psig applied pressure. Problems with this membrane included inherent fragility of the coated fibers and ulti-mately inconsistent performance results on test in the full scale element mode.

The vast majority of prior art sulfonated polyaryl- -~
1~ ether membranes has consisted of either (impractical) laboratory scale, thick dense films or asymmetric structures. These types of structures have been incapable of providing at least 15 GFD flux with 95% salt rejection at economical applied pressures (under 400 psig) as -~
required for RO applications. As a result, some researchers investigated thin film composite membrane designs, since this approach if carried out optimally yields maximum fluxes in conjunction with good salt rejections. This is consistent not only with theory but also with other known membrane structures in operation today. The limitations with prior art thin film composites rested with the techniques of fabrication and the polymer choices used.

SUMMARY OF THE INVENTION.

In contrast to the results of the prior art, it has now~been discovered that superior sulfonated polyarylether permselective membranes can be produced by deposition of particular solutions of these polymers directly onto ~ porous polymer substrates. This novel process is optimal-ly carried out without the need for precoating the porous ,~
` .
,.".,~

h i. .

-6- ~33232~
support or crosslinking the nascent thin film or coating.
It has further been found that many heretofore unexplored combinations of thin film barrier polymers and porous support polymers are possible, in fact - judicious combin-ations have proven optimal in the development of these thin film composite permselective membranes.

More specifically, thin film composite (or coated) reverse osmosis, ultrafiltration and microfiltration membranes are prepared from sulfonated polyarylethers with an ion exchange capacity (IEC) of 0.2 to 2.9 meq./gm., preferably 0.2 to 2.0 meq./gm., by a convenient one-step solution deposition onto porous polymer substrates. Polar solvents of medium to strong hydrogen-bonding ability, reasonable volatility, and low enough surface tension to wet the porous subst-ates are requi-ed. The optional addition of alkali metal salts, particularly lithium salts, to the coating formulation has no-mally been shown to dramatically increase water permeability of the resultant membranes. The deposition of said solutions of sulfonated polyarylethers (SP's) in the amount of 10-100 mg SP/ft2 of porous subst.ate followed by solvent evaporation yields composite or coated membranes with excellent permselectivity, toughness, and chemical resistance. These membranes are suitable for applications ranging f-om high pressure to low pressure reve-se osmosis (RO) (e.g., sea water to brackish wate- applications) to ultrafiltration (UF) and microfiltration (MF) applications including desalination, food and beve-age, pulp and paper, metal liquor, pha-maceutical, electronics, water soften-ing, medical, RO pret~eatment, and non-aqueous applica-tions.

' ~ ~ _ ~ ~ .

~ -7-~3~2~
The process consists of uniformly "coating" a dry or wet porous polymeric membrane substrate with a solution of a sulfonated polyarylether containing its sulfonic acid groups in the free acid form or salt form in a good solvent, or a mixture of liquids creating a good solvent, containing the requisite amount of a flux enhancing additive, then removing the bulk of the solvents by ;~
heating to yield a thin film composite or coated membrane. The solvent system, either as a singular liquid or as a mixture of liquids, is substantially polar, reasonably volatile, of low enough surface tension to wet the dry porous substrate, and, though possibly capable of swelling the porous substrate polymer, is not able to - -~
dissolve it. The selection of an improved solvent system lS for the sulfonated polyarylethers is an important and novel aspect of this invention. Effective coating has ;~
been demonstrated with a variety of controlled liquid application methods commonly employed in the coat-ing/converting industry. The result~nt sulfonated poly- -~ arylether thin film, while not chemically reacted with the porous substrate polymer, is remarkably well adhered to it and yields a tough, permanent and adherent layer that consistently survives both the chemical and mechanical rigors of realistic test conditions. Reverse osmosis ~5 membranes prepared in this way exhibit excellent fluxes ;~ and salt rejections, with uniquely high resistance to chlorine and other harsh cleansers commonly used for membrane separation applications. Ultrafiltration and microfiltration membranes prepared in this way have ~30 improved (1~ fouling resistance, (2) oxidative stability, and ~3) pore size distribution.
`~;:
:
:: ~

` -8-~33~2~

A high performance UF membrane having a porous aniso-tropic substrate polymer membrane and a thin coating of sulfonated polyarylether may also be produced according to the process of this invention.

Unlike the dense salt rejecting layer needed for RO, with the UF, MF and low pressure RO membranes the sulfonated polyarylether coating must itself be sufficiently porous to allow control of the desired molecular weight cutoff or salt rejection in the case of ~1~ low pressure RO membranes. The substrate in a coated UF
membrane of the invention will usually have a lower molecular weight cutoff (tighter) after coating than before, and accordingly, a suitable UF substrate will usually be selected from more "open" (higher MW cutoff) UF
~15 membranes than those which are desired to be produced.
; The molecular weight cutoff of the coated UF membranes of this invention can be controlled in one of two ways.
First, the molecular weight cutoff of the membrane can be controlled by adjusting the formulation of the coating 2n solution or adjusting the coating curing `cycle. Second, the molecular weight cutoff can be controlled by depositing the coating solution on ~F or I~F membranes with different pore size distributions. Using thiis latter approach, the coated UF membrane will normally exhibit a lower molecular weight cutoff (tighter membrane) than the - UF membrane support. One possible exception to this is when the coating is extremely light, such as when the purpose of the coating is to change only the hydrophilic nature o a polymeric substrate membrane. Sulfonated ~30 polyarylether sulfone or sulfonated polysulfone coating solutions can be deposited preferably on polysulfone or ? polyethersulfone UF membranes to provide a series of ~, -9- 133232:~

composite UF membranes with closely controlled molecular ;~
weight cutoffs.

Although ~e do not wish to be bound to any theory expressed herein, it can be postulated that the RO mem-S branes have a continuous film coating free of large pores to allow diffusion of water while rejecting salts whereas in UF and MF membranes the sulfonated polyarylether layer should only coat the surface of the membrane without substantially filling in, or coating over, the original pores in the polymeric membrane substrate. Although dir-ect proof of the correctness of this belief is lacking, it may be helpful in understanding the present invention to consider the RO-type membranes as true composite membranes with a continuous gel polymer film (barrier layer) at the surface of a porous membrane substrate, whereas the UF
; membranes may be thought of as being porous UF membranes with a light coating of the sulfonated polyarylether polymer conforming to the original geometry of the substrate (restricting, but not completely coating over 2~ the pores). As used herein, the term "composite" is mainly intended to relate to high pressure RO membranes of the invention (useful for desalinating sea water or the like) and the term "coated" relates mainly to low pressure RO, UF, and MF membranes of the invention, except where otherwise the terms are used in a more generic sense.

:~
~:
' ~ ' ' y1~

~ 33232~

DETAILED DESCRIPTION
It has now been found that thin film composite mem-branes with superior flux and re~ection properties can be fab~icated by a solution deposition process utilizing sulfonated polyarylether polyme~s containing a specific range of sulfonation content togethe~ with selected sol-vent systems and po-ous polymeric substrates.

The polya~ylethers useful in the practice of this invention are aromatic polyme-s lacking oxidatively un-10 stable linkages in the main chain and of sufficient molec-ula- weight for good film fo-ming behavio-. These belong to th~ee different basic st.uctu-al types as represented : by the fo~mulae below: r 15 I f 0 - R- 0 ~ S ~ 0 ~ R

: .
in which R is an a~omatic radical chosen f~om the follow-ing:

l ~ whe~e a = 50-200 and b = 0 O
CY~
-2 ~ ¢ ~ whe~e a = 50-200 and b = 0 c~

-3 ~ where a = 50-200 and b = 0 -4 ~ ~ where b/a = .1-50 ,~ . .

",, ~z. ~
. i.. ~ ,, ~. . , ~ , .

33232~
~:

_5 ~ where b/a = 1-20 -6 ~ where a = 50-200 and b = 0 . ''~

-7 Any combination of the above radicals and addition~
ally those not shown above that are derived from commen-cially avallable dihyd-oxy aromatic monomers capable of reactlng with a dihalodiphenyl sulfone monomer to fo~m a polya~ylethen sulfone.

II ~ 0- ~.- 0 ~ C ~ n = 50-200 1~

in which R is an aromatic radical chosen f~om the follow-ing:
-1 ~ -2 ~ -3 ~ o ~

~4 An~ combination of the above nadicals and alterna-~1; tive'~, those not shown above that a~e derlved f-om com-me-^ all~ available dihyd-ox~ aromatic monome~s capable o~
rea~ting with a dihalodiphenyl ketone monomer to fo~ a pol~aryle~;~en ketone.
i (S3Z)x Rl ~
III ~ n = 50-800 ~ J n ~

' .''~ ' '~ ~ '.. , :~' ~ ' j:
~i",,."~,~"i' ' ~ :. ' ~'' ' `
~ 2 3 2 ~

in which Rl and R2 are hydrogen, or alkyl radlcals, thus speciflcally: r ~ S3 ~ I O _ 3 ~ ~ (S03Z)x n S03Z refe~s to the sulfonlc acid group which is introduced into the fully synthesized polyme~ by methods commonly practiced for aromatic polyme~ sulfonations. The position of this g-oup, consistent with electophilic aromatic sub-stitution, ls p-edominantly ortho to the activating ethe-linkage for all polyme- types shown. Z may be hydrogen ln the case of f-ee acid, an alkali metal salt, or a nitrogen cont2ining salt de-ived from ammonia or amines. The value for x, which is the numbe~ of sulfonic acid residues pe-polyme~ repeat unit, will va-y for each polymer example and can be calculated f-om the ion e~change capacity (I.E.C. exp~essed as milliequivalent3 of sulfonic acid residues pe g~am of sulfonated polyme~). The latte- is dete~ined analytically f-om sulfu- content3 found with -0 eac~ sulfonated polyme-. The useful sulfonic acid content ~ange is found to be diffe-ent for e ch individual polymer type, but in general ls ve~y simila~ for polymers within the same structu~al fo~mula class as shown in Formulae I, II, and III.

The ave-age I~C requi-ement fo~ memb~anes of this ¦ invention becomes g~eate- as one compares polyme~s of class I through III, and this is believed to be due to cor-esponding dec~eases in hyd~ophilicity in the unsulfon-ated starting polyme~. Thus while polyarylethe- sulfones of class I a-e effective with an I~C -ange of about 0.3 to 2.2 meq./g-am, polyphenylene oxides of class III must usually be in the -ange of about 1.3 to 2.9 meq./g-am.
' ~
, :~
l ;~
, :
I ~

,t 5.' ~ . . ' ':: .. ' ' 3~2325 There are two requirements of the polymers used in the present invention with respect to sulfonic acid content.
The first requirement is that these polymers be soluble in the preferred solvents of the invention and yet not be appreciably soluble in a membrane test feed - such as saline water. The ranges of IEC given above cor-espond to the minimum degree of sulfonation for solubility in the preferred solvents and maximum allowable degree of sulfon-ation for stability in aqueous environments, respectively.
It is pointed out that as these polymers become gradually more sulfonated they progress from solubility in solvents of low to moderate polarity and hydrogen bonding ability to ones with moderate to high values, and ultimately to solubility in water.
:
The second requirement is that the IEC be that which yields the best combination of membrane flux and rejection pe-formance for the particular separation application.
Fortunately, this is nearly always found to be near the ~;~ center of the broad ranges specified. It is gene~ally ~20 found in practice that higher membrane IEC values lead to higher water permeability with proportionately higher solute passage and conversely, lower membrane IEC values lead to lower water permeability with proportionately lower solute passage (higher rejection). Thus it is pos-sible to tailor the performance of composite RO membranes ; ;~ by optimizing the degree of sulfonation of the polyaryl-~ ether thin film polymer to narrow ranges of IEC.
, ~; The c_iteria for the solvent system employed in the coating process is that it must be a good solvent for the ~30~ sulfonated palyarylether, a nonsolvent for the porous subst-ate, of low enough surface tension to wet the porous , ~

~ .

~ s , ., .. ,.. ,, " .,. ,~
"~
', ~

f -`

-14- ~332325 substrate, and volatile enough to be removed by gentle heating. The fact that the sulfonated polyarylethers and the porous substrate polymers may be very similar chemically may limit the number of effective solvents, though investigation of solubility parameters for both polymers has uncovered several ~iable types.

The present invention resides largely in the finding that the membrane properties of sulfonated polya-ylethers can be surprisingly enhanced when applied to a subst_ate from a particular novel, potentiating solvent system. The membranes made according to this invention thus have a combination of improved flux/rejection when compared to sulfonated polyarylether membranes prepared heretofore using the conventional solvents and solvent systems.
Accordingly, as used herein and in the appended claims, the expression "potentiating solvent system" is intended to describe novel solvents and solvent mixtures heretofore unknown for sulfonated polyarylether coatings and which unexpectedly produce membranes of improved rejection and/or flux qualities.

Foremost among the potentiating solvent systems are those comprising significant amounts of formic acid, that is about 10% to 100% formic acid, with the remainder being ~` cosolvents and flux-enhancing additives, etc. More ~;25 preferred are solvent mixtures containing at least 20~
formic acid. Additional solvents and solvent mixtures found to enhance sulfonated polyarylethe membrane properties include fo-mamide and formamide-containing mixtures such as formamide/2-methoxyethanol and formamide/formic acid. Solvents which ordinarily would not be useful for casting sulfonated polyarylethers on a .
~ ~ .

::
`:~
"

'':~
~; ;~, , ;, t ` -15- ~33~32~

conventional polymeric substrate membrane because of either their tendency to dissolve the substrate, or their inability to dissolve the sulfonated polyarylether because of poor solvent ability, such as acetone and acetonitrile respectively, can be used in combinations with other solvents, such as water, to form satisfactory membranes by the process of the invention. Thus, mixtures of acetonitrile/water (95%/5%) and acetone/water (75%/25%) have been shown to produce acceptable composite membranes on open polymeric membrane substrates.

:
:
~ , ~ .

,~
.~
~ ' ., '`
, ,:
;:
:~ `

~`:
:~
~:

~ ~ .

~ . ; I _ 'r~

.'5~
~:. `; ~

.~ ~

-16- 133232~

The solubillty parameter approach, partlcularly the three-dimensional system proposed by Hansen has been used effectlvely in predicting liquid miscibility and polymer solubility ln Rome instances. mis three-dimensional system assumes that the total cohe~ive energy which holds a liquid together (E) can be divlded into contributions from dispe-sion (London) forces, Ed; permanent dipole-dipole forces, Ep; and hydrogen bonding forces Eh. Han-sen's equatlon ls as follows:
~ 2 j~ 2 ( 2 r 2 t _ ~ d 0 P ~ h where 0 d 5 (Ed/V ~2 p = (Ep/V ~2 S h = (Eh/V h2 :~
and V = molar volume of the solvent.
Table I below shows the total and component solubil-lty parameters as well as othe- physical data fo~ two ~ul~onated polyarylethersl a porous substrate polymer (polysulfone)~ and solvent systems which are compatible ~20 wlth the porous sub~trate polymer. The primary dlffe~en-` ~ ces between sulfonated polymers I-2 and I-~ and polysul-foneiparameter Yalues are found with the ~ h and ~ p components due to sulfonlc acid functionality and de-.
.,~

..

i ~.,: ~

~;' :` ~
-17- 133232'~

, creased hyd~ocarbon character of the former. Wate~ a ~! non-solvent which has a very high value for the ~ h component, ls added to other non-solvents (for sulfonated polyaryl~ethers) such as acetonitrile and acetone, whose ~ h components are small (3.0 and 3.4, respect-ively) in order to produce good solvents.

Experiments have shown that the high boiling forma-mide solvent is incapable of producing thin fllm RO compo-site membranes with high salt re~ection, and this is be-lieved to be due to a combination of excessive su-face tension and high boiling point of this liquid. Formamide can be employed however, in smaller amounts with other solvents. 2-methoxyethanol, a weaker SP solvent, often requires the addition of 5% or more formamide or at least 20% for-mic acid for adequate solvency and even greater percent-ages of the latter for high R0 membrane salt re~e~tion.
Addition of water to this solvent doesn't lmp~ove SP s~lu-billty,~p~Qsumably due to an imbalance in resultant ~, p and ~ h component values. Addition of inc-ePsing water contents to such systems such as formic acld, acetonitrile, and acetone produce a trade-off ln ~esultant memb~ane perfonmance, amounting to higher fluxes and ~ proportionately lowe~ salt re~ections. It has beQn found ¦~ that, in general, any strongly-hyd-ogen bonding s~bstance ¦ 25 when added to the coating solution produces this behavior. This allows another dimension ln tailo-ing formulatlons to meet various desl~ed water permeaDlllty-permselectivlty goals.
~:
It is important to note that the solvent systems for SP's shown in Table I have been obsenved to swell a porous polysulfone R0 subst.ate polymer wlthout actually damaging ,'~
' , _ ~,:
, "
'`,''~
., '.-. - :- ..
,' ~

~ -18-1 ~232~
it, and although this invention isn't limited by theory, it is theorized that this behavior leads to desirable entanglement of the thin film and porous substrate polymers in the composite RO membranes of the invention.
It is further deduced that this expected phenomenon accounts for the remarkable adhesion and durability of these thin film composites. Formic acid has been found to be an extremely effective solvent for the practice of this invention and is thus preferred. The power of this solvent is believed to be the result of not only a good solubility parameter profile, but also the result of its small size and ionizing ability. Also advantageous are (1) the ability of this solvent to swell a porous polymeric subst~ate without harming it and (2) relative lS ease of evaporation (b.p. =101C).
j,,~
A demonstration of the swelling action of formic acid -~ on typical anisot opic polysulfone substrates used for SP
membrane fabrication is illustrated below in Tables A and B. Included in these tables are comparisons with 2-methoxyethanol (ethylene glycol monomethyl ethe-) which `~ has been used by others in SP membrane fabrication but which by itself is not as effective as formic acid. As seen in Table A, formic acid yielded 50% greater subst_ate water flux than the control (isopropanol/water) treatment ~25 whereas 2-methoxyethanol yielded vi.tually the same water flux as the control. Table B indicates even more dramatically the differences between solvents. Here polysulfone membrane substrates were coated with solvents and dried using the same technique employed during SP
~30 membrane fabrication. Testing of these relatively hydropbobic polysulfone substrates was performed by first rewetting briefly at 400 psig, then lowe ing to 223 psig :~
::~
, ...
~, ~ ~ t ~ `

:-.,` '~

~ j -19- 133232~
~^ ' at which pressure water fluxes were taken. The formic acid coated example yielded 23% greater water flux than the control whereas the 2-methoxyethanol case produced approximately half the control flux.

Table A

Water Flux Rates of Polysulfone Membrane Substrates Soaked in Different Solvents Then Finally Equilibrated in Water , .-. Relative Flux Rate IO Solvent Treatment @ 55 psiq 50:50Isopropanol/Water (Control) lOO
Formic Acid 150 2-methoxyethanol 95.6 :~
~ Table B
::~
`~15Water Flux Rates of Polysulfone Membrane Subst-ates Surface Coated with Diffe_ent Solvents, D-ied 3 ~:Minutes @ 55C, Rewet with Water @ 400 psig Treatment Relative Flux Rate Q 223 psiq None tcontrol) 100 ~';0 Formic Acid 123 2-methoxyethanol 47.2 ~ i .
There-is, in addition to the strongly hyd-ogen bond-ing liquids just mentioned, a wide variety of compounds, consisting of alkali metal salts, organic acids, and or-~, ~, 5 ~ ',. `' i ~ .' " `''' ~ `;~ ~ ~ - ~ . ' :
i',~

' 20 ~33232~

ganic nitrogen containing compounds which are extremely potent in their ability to increase water permeability of the SP membranes. While many such compounds are effect-ive, lithium chloride is preferred due to its potency in maximizing water flux of SP membranes at very little pen-alty to salt rejection. Although this invention is not limited by any theory, it is theoretically possible that lithium ion with its large hydration per size ratio is able to retain maximal water contents whether as the 10 counter ion of the polymer sulfonic acid and/or simply as bulk salt (hygroscopic) residue distributed throughout the thin film of an RO composite membrane. The amount of lithium chloride incorporated into the coating solution formulation is dependent on the particular SP polymer 15 being used, which is ultimately dictated by the desired combination of membrane flux and rejection performance.
Solubility parameter data for some typical solvents , and non-solvents for the SP's used by the present invention are set forth in the following Table.
,~ .
~, ,~

:
: .

~4;,'~'.. ~.. ' ' ' `' .'' '''~ 1.' ' . '',; . ~ `~ ,, !~
;~ , .. ", .~ . , ., ., ~ , ~ ~ ~ ,. .
.~ ,~. ,,, . , ", . ,~ ~ ~ - :

--21-- . , O ~ ~D O O~

~V V ~ U~ U~C~
I I I Ii 3 r-1 N t-- I I I
O ~ ~ _~ ~ ~
CC~ U~ L~ N
aa r-- N
~_1 ~v _I C-- --I _I L~ o L~
_ o ~J C ~ O
~, o ~ I I N ~ I C

C) V _~ L'~ N N J 0 O ~ N L~ L~ CO :~ 0:1 . ~ ~ t~ ~ ~ N
_ --I O
r c~ .
~U ~ _~ N C~ ~ ~ Lr~ ~ ~ 11 ~1 ~i _ L~ ~ t~ tY ~ ~ N N . .
~: -o~o ~
_ I --~ J ~ '-I ~-- J
_ ~ ~ C ~ J C C ~ ~) ~ L'~
~0 ~, ~N 3~ N
. ~ ~, 3 3 CO ~r) Lr~CO L'~ CO CCi ~a) 5 V g ~ ~ ~ :
, N CO ~ 3 t--CO Ll~ 9~ V ~ Q.
~ ~1 --O ~ v ~ ~
J t-- h 3 ~D ~D O ~ ~D L'~ ~D h V ~ t-- _I
~C~ co co ~ co t~ t-- r-- ~ _I
~' ~ ~ ~ `D
'~ ~ OD CO ~ N -~ J O. CO ~ æ
:~ ~ bD N o~ N N ~N 0~ V ~,, ~ r ~ ~ r~ ~ ~ O

U I I1 9~ ~ cq C4 S~ t3 ~ ~ ~ .
. ~ d ~, v ~! I V i~
~ ,,_ ~ 1~ a~ ~

2 ~ 4s v. ~ ~ ~ V ~1 ~ ~P

2 ~ ~ ~ O D~ ~ bi ~ 0 ~ ~ ~ e ~ ~
~0,~ pO,~ ~ ~ ~ ~ N ZO :~ ~ d --I N ~ 3 Lt~

It has been demonstrated that deposition of the SP
solution onto the porous polymer substrate can be accom-plished by a variety of controlled coating processes available from the converting indust_y. The amount of SP
polymer deposited on the porous subst-ate should be as little as is needed for thin film st~ength and continuity in order to yield maximal water penmeation with good sol-ute rejections. In practice, this can be from 10 to 100 mg of SP polymer per square foot of porous substrate al-though the preferred loading is from 20 to 50 mg per square foot. Estimated thin film thickness for this load-ing is on the order of 2150-5400 A (.215-.540 microns).
:~
Laboratory scale coating of porous substrates to form RO composites has been performed conveniently by brushing on the SP solution, whereas continuous (moving web) coat-ing (pilot scale production) has been demonstrated on more sophisticated machine processes, such as single and double roll bead, rod, and spray coaters. Subsequent to the solution deposition, a solvent removal step must be per-formed which yields an essentially dry, thin film com-posite. This consists of passing ai- at 0-120C tempera-ture and velocity of 0-3,000 feet~minute across the nas-cent thin film surface for a period of a few seconds to several minutes or even several hours. The solvent remov-al step is a critical one with respect to the final trans-` port properties of the thin film composite RO membrane.The removal process is keyed to time, temperature, and velocity of the air applied, with g-eater values of these conditions yielding lower water permeability-higher perm-~30 selectivity membranes, and lesser values yielding con-versely higher water permeability/lower permselectivity ` ~ membranes. For RO membranes, the preferred temperature ~ ~,~
': ~

J~ ` ~
i i -2~ 13~232~

and time range is ~0-80C and 0.5-10 minutes, respective-ly. The forced air velocity is less critical but can be used advantageously to compensate to some degree for less-er temperatures and/or times applied. It must be men-S tioned that other methods of heating such as infrared etc.
with or without forced air have been investigated and found to be acceptable but less desirable than the pre-ferred method.

Optional post treatments consisting of organic sol-vents and aqueous mixtures with or without thermal treat-ments can be given to the finished composite RO membrane for the purpose of improving permeability via swelling and/or hydration of the thin film. The post treatment may alternatively include further stabilization or "tight-lS ening" of the thin film polymer through ionic crosslinking or salt formation using solutions of multivalent metal salts, or through basic nitrogen containing compounds.
Thermal t-eatments in aqueous media may be applied for the purpose of reordering the permselective barrier and thus further affecting membrane t_ansport behavior. It should ` be understood that these post treatments are not normally required or performed but may be pe-formed as an optional refinement step to achieve specific desi-ed membrane per-formance.
:-The choice in porous substrate is governed by much of the same criteria applied to the SP polymer except that resistance to solvents for the subst~ate is requi~ed.
Again it is preferable to utilize polyme~s with a high degree of resistance to chemical degradation-especially by ~30 oxidizing species such as chlorine, damage caused by wide ranges of pH, bacteria and enzymes, flow and creep under ~, : ~.
:~

~, i ~
~`'~: .'~ ~ '' ` ~'~'""' ~ , .,.'.'.,.,-:
: :~

i.
-24 1~32~2~

pressure, and certain organic solvents. As with any polymer application, tradeoffs in these properties exist and only careful consideration dictates the best choice of a specific porous substrate polymer. Examples of porous substrate polymers useful by the invention include but are not limited to: polyarylether sulfones, polyaryletherke-tones, polyphenylene ethers, polyphenylene thioethers, polyimides, polyetherimides, polybenzimidazoles, polyes-ters, polyvinylidene fluoride, polychloroethers, polycar-bonates, polystyrene, polyvinylchlorides, polyacrylonit-rile and various copolymers of the last three types, etc. Inorganic porous materials such as ceramics and metals may also be used as substrates in some specific instances. The preferred polymers for thin film composite lS membranes and for the coated UF membranes of this invention are polysulfone and polyarylether sulfones.
These porous substrates may have surface pores of a size ; range of .001 - .5 micron (1-500 nm), though the preferred range for RO and UF applications is from .001 - .03 microns (1-30 nm). The preferred sizes of pores are ideally small enough to prevent intrusion of the thin film SP solution, which would result in low permeability, and at the same time are adequately large for insignificant contribution to hydrodynamic resistance. A further problem ~25 with excessively large pores in RO applications is the inability to support the thin film polymer of a composite RO membrane under applied pressures resulting in losses to permselectivity.

A critical relationship has been found to exist be-tween the particular SP and porous substrate combination ~ and the resultant composite membrane performance. It has 9 been found that membranes made from porous polymer ,,; ,,. ~

,,.,. .~

.
~ -25- ~332323 supports containing substantially only ether and sulfone linkages have consistently lower flux and higher salt rejection (are "tighter") than ones made from porous supports containing additional alkyl linkages. Whether this is due to pore size differences, surface energy differences (and thus wetability), or other phenomena between the more hydrophilic ether-sulfone and the less hydrophilic alkyl-ether-sulfone types of aromatic porous substrate polymers isn't known at this time.
Nevertheless, an important empirical relationship has been uncovered and put to use in this invention by combining the appropriate SP and porous substrate polymer to yield the best thin film composite membranes and coated UF and MF membranes.

The above mentioned porous substrates useful by this invention may be in flat sheet form with or without an additional porous polymeric supportive sheet (such as those car-ier fabrics commonly used as moving webs in the flatsheet RO or UF membrane industry), in hollow fiber form, o- in tubular form.
. .
Spi al wound flatsheet, ordinary flatsheet, hollow fiber, or tubular modules prepared from thin film compos-ite membranes of this invention are useful for separations by reverse osmosis, microfiltration and ultrafiltration.
~25 Utility of these membranes includes both purification of water through removal of salts, organic compounds, viruses, bacteria, colloidal substances, sediments, etc.
as well as recovery or concentration of valuable substances pertaining to dairy, fermentation, paper &
pulp, fruit juice, electroplating, mining, pharmaceutical, electronics, painting and chemical industries.
`
'. ..
::
:~

~_~.~,.~.~.. ,.,,.. ,., . ~ . _ ~,, ~

~ ,:j ~

-26- 13~232~

Particularly advantageous is the utilization of the above membranes for separation processes in which high tolerance to oxidizing agents such as chlorine and/or extremes of pH
are desired.

The following Examples serve to further illustrate the invention but should not be construed as in anyway limiting the broader aspects thereof.

:
An 18% (by weight) solution of polysulfone ~Udel 1~ P3500, Union Carbide) in 75% N,N'-dimethyl formamide and 7~ bis-(2-methoxyethyl) ether was cast on a paper-like porous sheet then gelled in a water bath to produce a porous substrate. This porous substrate was subsequently dried at 105C for 5 minutes and brush coated evenly with 1~ a .25% (wt./vol.) solution of sulfonated polyaryether (formula I-5 above - 1:5 sulfonated ~ ictrex - ICI
Americas, IEC = 1.29. meq/gm. sodium salt form) containing 0.14% lithium chloride in formic acid (90% solution).
After forced air drying at 55C for 3 minutes, samples of n this membrane were tested for reverse osmosis properties on a 2000ppm, pH8 sodium chloride feed at 200 psig (NDP). The performance after 1 hour for a mean of 6 samples was: 16.8 GFD flux and 96.7% salt rejection.
* Trademark -~ j ~' .

,., :~.. ~ i, .. .... ... . .

~:.
~3 ~ -27- ~3~23~3 A porous substrate was prepared and coated as in Example 1 except this time with a 0.6% SP-containing solution with no additives in a solvent mixture of 95%
acetonitrile/s% water. After the same drying treatment as Example l, samples of membrane tested on a 5000 ppm, pH6 sodium chloride feed at 343 psig NDP for l-hour gave 15.5 GFD flux and 88.9% salt rejection.

An RO membrane was prepared as in Example 2 except 0.25% SP was used and 7% of the acetonitrile in the coat-ing solution was replaced by formic acid. Samples of this membrane gave a mean (3 samples) performance of a 9.6 GFD
flux and 96.9~ rejection when tested for l-hour on a pH8, 2000 ppm sodium chloride feed at 377 psig NDP.

, A membrane was prepared and tested as in Example 2 except the coating solution consisted of 75% acetone/25%
water and gave a mean (3 samples) performance after 1 hour ; 20 of 24.7 GFD flux and 80.9% salt rejection.

A membrane was prepared and tested as in Example 2 except 0.75% SP was used in an 80% 2-methoxyethanol/20~
formic acid coating solution and gave a mean (3 samples) performance after l hour of 27.5 GFD flux and 89.4% salt rejection.
~:
,~
,~

``

- ~
-28- ~33232~

A membrane was prepared and tested as in Example 5 except that a 94% 2-methoxyethanol/6% formamide coating solvent system was employed and this membrane yielded 136 GFD flux and 38.0~ salt rejection after 1 hour (mean of 3 samples).

A membrane was prepared and tested as in Example 1 except that the coating solvent was 100% formamide. Mean performance (2 samples) after 1 hour was 127 GFD and 12%
salt rejection.

A 21% (by weight) solution of polyimide (UpJohn 2080) was formulated as described in U.S. Patent 4,307,135, cast on a pape- like porous polymer sheet, then gelled in a water bath to produce a porous subst~ate. After saturat-ing this substrate in a 20% solution of glyce-ine in meth-anol for 8 minutes, this subst ate was coated and dried as in Example 1. Testing on a pH8, 2000 ppm sodium chloride ~20 feed at 377 psig NDP gave 1S.7 GFD flux and 99.5% salt rejection afte~ 1 hour (mean of-2 samples).

f EXAMPLE 9 A porous support was prepared as in Example 1 except ~ ~ that Victrex polyethersulfone (300P ICI Americas) was used S for the support. When coated and tested as in Example 1, -29- 1332~2~

resultant membrane performance was 6.5 GFD flux and 98.0%
salt rejection (mean of 4 samples).

A porous support was prepared as in Example 1 except that Radel~ polyphenylsulfone (A-400, Union Carbide) was used. When coated and tested as in Example 1, resultant membrane performance was 4.8 GFD flux and 97.6% salt re-jection (mean of 3 samples).
:

~1~ A porous support was prepared as in Example 1 except that a 90~ Udel polysulfone/10~ Victrex polyethersulfone mixture was used. This substrate was coated and tested as in Example 1 except that 377 psig NDP was used in testing.
Membrane performance after l-hour was 23.8 GFD flux and ~15 98.0% salt rejection (mean of 4 samples).
~ .
; EXAMPLE 12 :~' A membrane was prepared as in Example 1 except that formula I-l, above (IEC = approximately 1.0 meq./gm., sodium salt form) was used in conjunction with 0.07%
lithium chloride. Membrane performance, when tested as in Example 1, was 13.6 GFD flux and 96.6% salt rejection (mean of 3 samples).

~:~
A membrane was prepared and tested as in Example 12 except that the porous substrate used was that in Example ~, -:

r r, - ~

`~ `:

~ --30-- ~ 3 3 2 3 2 ~
::j 9. Membrane performance was 6.0 GFD flux and 97.9% salt rejection (mean of 4 samples).

A membrane was prepared and tested as in Example 1 except that formula I-2 SP, above (IEC - 0.86 meq./gm., sodium salt form) was used and at a concentration of 0.5%. Membrane performance was 17.4 GFD flux and 90.8%
salt rejection (mean of 3 samples).

A membrane was prepared and tested as in Example 14 except the coating solution was diluted in half and the porous substrate of Example 9 was used. Membrane perfor-mance was 19.5 GFD flux and 93.9% salt rejection (mean of ~ ~
2 samples). ~ -,: ~
lS EXAMPLE 16 A membrane was prepared and tested as in Example 1 except that a solution of 0.5% of formula I-4 SP, above ; (IEC = 1.9 meq./gm., sodium salt form) without additive was used. Membrane performance was 234 GFD flux and 20% ~-~
salt rejection (mean of 3 samples).

k EXAMPLE 17 : :'- ~
A membrane was prepared and tested as in Example 16 except that the IEC was 1.29 meq./gm. and the porous substrate used was that of Example 9. Membrane perfor-~25 mance was 8.1 GFD flux and 93.4% salt rejection (mean of 2 samples).
, ' :

~ i~

' -31- ~332~2~

A membrane was prepared and tested as in Example 17 except that 0.28% lithium chloride was included in the coating solution. Membrane performance was 60.1 GFD flux and 59.4% salt rejection (mean of 2 samples).

A membrane was prepa-ed and tested as in Example 17 except that the coating solution was diluted in half and the porous substrate was that used in Example 10. Mem-brane performance was 7.6 GFD and 93.7% salt rejection (mean of 2 samples).

:
A memb~ane was prepared and tested as in Example 16 except that the coating solution contained formula II SP, above (PEEK~ 450P, ICI Ame icas, IEC unknown, sodium salt form without additive!. Membrane performance was 15.3 GFD
flux and 92.8% salt rejection (mean of 2 samples).

~:
A membrane was prepa.ed and tested as in Example 16 except that formula III-l SP, above (IEC unknown, free acid form) was used. Membrane perfo~mance for this case was 17.1 GFD flux and 90.5% salt rejection (mean of 2 sam-ples).
~:

:
' ~ '.

~5 `. ~ -~f: `
`

-32- 13323~3 i;~

A porous polysulfone substrate prepared as in Example 1 was coated on a continuous processing machine by spraying a 0.22% by wt. solution of sulfonated polyethersulfone (IEC = 1.36 meq./gm.) in 90:10 formic acid: water containing 0.2% lithium chloride. This coating solution, which was applied at a loading of 13 g/ft2 was evaporated in a forced air oven initially at 70C for a period of 35 seconds then at 80C for 2.5 minutes to yield a composite membrane. Samples of this membrane when tested on a 2200 ppm sodium chloride feed (pH 8.2) at 200 psig NDP yielded 14.9 GFD flux and 97.0%
rejection after 24 hours.

Demonstration of reverse osmosis memb~ane performance stability on test with and without added 6 ppm active chlorine at pH 8.2 has been car~ied out using two inch spiral-wound elements of membrane ( 4 ft2 area) made by a continuous machine, based on the fabrication method of Example 1. After 420 hours of equilibration testing on natural pH 8.2 San Diego tapwater ( 600 ppm total dis-solved solids-"TDS") at 200 psig NDP, perfoemance was 7.5 ;~ GFD flux and 97.5% rejection (TDS). After this point, 6 ppm average free chlorine was introduced and kept at this level while maintaining all of the original test condi-tions for a period of 550 hours. Flux and rejection per-formance remained vi-tually unchanged at 7.4 GFD flux and ~-~
98.1% rejection. Not only has stability on a live feed been demonstrated, but also excellent tolerance to a com-monly used water disinfection chemical - chlorine.
' ~ :

.~ . ~. .. , . ~, ~: ~. -: -~J`~ ~ ."~

3 _33_ ~33232~

Another two-inch spiral wound membrane element very similar to that described in Example 23 was RO tested on a pH 8.3 live San Diego tapwater feed containing 670 ppm TDS, at 392 psig NDP. Equilibrated performance on this feed consisted of 15.1 GFD flux and 98.1% rejection. This membrane element was then subjected to a variety of ag-gressive aqueous cleaning agents, many of which are used in actual RO membrane applications. Table 2 below lists the sequence of stepwise closed-loop recirculations of the various chemical agents. After ali of these treatments, many of which would be too harsh for other commercially important membranes, including cellulose acetates, poly-etherureas, and polyamides, the tested membrane element I5 yielded virtually the original performance, that is, 18.4%
GFD flux and 98.0% rejection ::

:, Time ~circulated Aaent ConcentrationpH (Hours) Sodium Hydroxide and Sodium Chloride O.lN 12.8 Citric Acid 1% 2.4 2.3 Urea 30% 8.2 4.1 ~25 Sodium dodecyl sulfate 1000 ppm 7.5 1.6 -~ Sodium hypochlorite 115 ppm 9.1 1.5 Sodium bisulfite 1000 ppm 4.5 3.1 Oxalic Acid 1% 1.8 EDrA and sodium tripolyphosphate 1~ 9.8 .

. . ~

l ~
. ~ :. . .

-34- 133232~

In ultrafiltration applications, one very common membrane form for industrial and laboratory uses is the hollow fiber. Hollow fiber membranes are mounted, usually in cartridges, with the open ends of the fibers potted at each end in an adhesive plug sealed to the cylindrical cartridge walls at either end thereof. The membrane bar-rier layer or "skin" is usually on the inside of the hollow fiber or "lumen". The process feed usually enters one end of the hollow fiber and a concent ated solution exits at the other end. Pe_meate, i.e., the liquid which passes through the membrane normally exits the cartridge ;~
from one or mo.e ports in the cartridge shell.
~ ~:
To demonst-ate the effects of coating ultrafiltration membranes with a sulfonated polyarylethe- sulfone polyme~
industrial grade membranes in cartridge form we-e treated. The fiber types which were coated included polysulfone ultrafiltration products having molecular -weight cut-offs ranging from 1,000 to lO0,000 and with fiber internal diameters from 0.02 inches to 0.06 -~;~
inches. Among the polymers used for the substrate fibers we-e Udel~ and Radel~ (Union Carbide Corp.) and Victrex (ICI).
:~, .,; .:-To coat the hollow fibers, the cart~idge containing the potted membranes was clamped vertically and the 25~ coat~ing solution was applied to the interior of the fibers ~-only. The polymer solution was held in contact with the ~--membrane surface for a short period usually about 15 seconds, and then allowed to drain. Afte- the hollow fiber was coated with the polymer solution, the cartridge ~30~ was placed on a curing fixture where heated dry air was forced through the lumen of the fibers. Heated air can ::

,,`j ~

_35_ ~3~2~

also be blown onto the exterior of the fibers. The curing parameters which can be varied are air temperature, air flow rate and drying time. The temperature of the heated air in the fiber lumen can range from 30C to 100C, s although 48-58C is preferableO The air flow rate can range from 0.25 CFM to 12 CFM and is highly dependent on air temperature and time.
., The following specific examples serve to teach the formation of coatings or films on commercially available 1~ hollow fiber membranes to produce UF, MF or low pressure RO membranes. The SP's were prepared by processes either illustrated above or otherwise well known to the art heretofore. A common method of sulfonation of polysulfones may be found in Coplan U. S. Patent ~o.
lS 4,413,106, issued November 1, 1983.

A reverse osmosis membrane was formed on the internal ; surface of a dried hollow fiber support using a sulfonated polysulfone polymer having an IEC of 1.29 - 1.4 meq./gm.
2~ This was accomplished by exposing the interior of the support fiber to a solution containing 0.50~ of the sulfonated polymer, 0.56% LiCl, and 99.36% formic acid for 30 seconds. Heated, dry air was thereafter passed through the lumen of the fibers at 50C and a flow rate of .17 CFM
~5 for 10 minutes. The performance of this composite membrane was 15 GFD flux and 75% salt rejection when ; tested on a feed of 5000 ppm NaCl and 240 psi NDP.

A reverse osmosis membrane was formed following the ~30 same procedur- described in Example 25 except that a wet ~ ~ .
, ~

~ l ~ ~
, ,~
~, .~

~ -36- ~ 33232~ ~

support was used. The performance of this composite membrane at the same test conditions was 200 GFD flux and 25% salt rejection.
, ~5 Ultrafiltration hollow fiber membrane substrates with various pore sizes (molecular weight cut off "MWCO") were coated with the sulfonated polyethersulfone described in Example 25 to yield ultrafiltration membranes with lower MWCO than the substrate fiber. The resulting test data are shown below in Table 3.

Coated membranes were prepared using the same procedure as Example 25 except the support was partially wet. A reverse osmosis membrane coating was formed using this polymer. The performance of this composite membrane at the same test conditions as Example 25 was 80 GFD flux ~-and 50% salt rejection.

EXAMPLE 29 ~-' :'. ~:
A reverse osmosis membrane was formed using the same ~2~ proc~edure described in Example 25 except the cure time was '~ increased to 20 minutes. The performance of this composite membrane at the same test conditions as Example - ~ 25 was 9 GFD flux and 81~ salt rejection.
1~
Several common proprietary commercial grades of 25 ~ polyarylether polymers useful by the present invention are ~ ~illustrAted by the followlng formul~e:

~' ~

I ~

~ ~3323~i~

~ -~ li' ~
¦ Udel (t-ademark of l 5 Union Carbide) ~ ~
n m Radel (trademark of Union Carbide) ~-~ S ~\~
lS n ::
Vict~ex (t-ademark of I.C.I.) , Polyarylether polymers of the above types were sulfonated and used to coat various UF membrane substrates. A summary of the performance data f_om a large numbe- of individual tests using Victrex 600 P as the coating polymer on a Udel polysulfone UF membrane :~

~, j~
,~ ",-:, . ,,:,~ ~ .: ~:; .,.,,-~ . ~.... , :
,,~

, .. .,~
, ;; '~
, -38- ~ 33232~

substrate are presented below in Table 3. The composition - of the coating solution formulation which was primarily ~1 used was:
- Sulfonated polyarylether sulfone0.50%
Lithium chloride 0.14%
Formic acid (90%) 99.36%

Table 3 Polvsulfone Hollow Fiber Membrane (Udel) Perfo ~ nce Data ~25 D-s-i-a-~ Before and After ~einq Coated with Sulfonated Polvethe~sulfone (Victrex 600P) to Fo~m Coated UF ~rane BEFORE OOATING AFTER COATING
DE~N 5000 * DEX~UN 5000 POL~ULPONE W~TER SOLUTION W~TER SOLUTION
S ~ ~E FLUX FLUX REJECTION FLUX FLUX REJECTION
(MW CUTOFF) (GFD) (GFD) (%)(GFD) (GFD) (%) 2,000 46.3 38.0 46.0 12.7 10.5 80.2 10,0n0* 102.4 95.5 22.2 44.3 39.1 73.5 10,000 140.0 10~.2 28.2 63.6 59.5 64.2 30,000 88.2 77.8 38.6 73.5 61.4 58.2 50,000 86.5 74.4 25.5 93.4 77.8 48.1 100,000 239.0 170.0 9.1 275.0 180.0 13.3 ~cationic ~olyelectrolyte coated The above ~able shows the flux and -ejection of the 2S base polysulfone UF hollow fibe~ membranes before and , after being coated with sulfonated polyarylethe~ sulfone . to form the coated UF membrane. These data illustrate that in every case the coated fiber or membrane exhibits a lower moIecular weight cutoff (tighter memb_ane) than the ~0 base UF polysulfone membrane by vi-tue of the linear polysaccha:ide Dext-an 5000 molecule rejection data. The coated UF membrane exhibits highe~ rejection data for the Dext-an 5000 molecule than the uncoated o- base polysulfone UF membrane. These data also establish one of 3~ the primary advantages of this invention. For example, an Trademark 1~
I

~ .
~, ~
~,~

! . . ' .

~,1,", ~

~39~ ~3~2~
original 50,000 MW cutoff membrane can be converted into a modified membrane that exhibits a lower molecular weight cutoff (% rejection) and higher flux (both water and solution) than a corresponding uncoated 2,000 MW UF
membran~ (see data of Table 3, above). Second, the coated UF membrane of this invention would also possess a higher degree of hydrophilicity than the base polysulfone UF
membrane made from Udel polymer.

Another approach that can be used to control the molecular weight cutoff of the UF membrane is to modify the coating curing procedure. Table 4 below, shows how the same base UF polysulfone hollow fiber membrane that is coated with a sulfonated polyarylethe- sulfone can produce a series of coated UF membranes with a range of molecular weight cutoffs by merely modifying the coating curing procedure. The advantage of this technical approach is ~hat the exact membrane rejection characteristics can be tailored to a specific application.

Sulfonated Polyethe_sulfone (Victrex 600) Coatinq Curinq Time On the Perfonmance Data of UF Coated Membrane ( 25 p.s.i.g.) CURING TIME W~TER FLUX FL~X REJECTION
(MIN) (GFD) _(GFD) (%) 0* 105.5 84.7 15.6 127.1 100.3 29.1 128.0 96.8 28.8 76.1 57.1 58.0 51.9 39.8 68.3 , : 30 10 (+18 hrs. at 50C) 38.0 31.1 76.4 10 (minutes of air flow 98.6 76.1 36.9 in each di~ection -20 minute total) *substrate of Udel polysulfone UF membrane :
~ .
~ ,,.

:.' ,~.i.," "~
~; ~
~ L .~
,:, i,~

i ~40- 1 3 3 2 3 2 ~
The UF coated membranes of this invention that are prepared by the methods discussed above exhibi~ the ~ollowing advantages:

(1) the coated membranes are more hydrophilic than the base UF polysulfone membrane.
:~.
(2) the coated membranes are strongly negatively charged.

(3) the SO3H moiety is highly reactive and can be -easily converted to chemical moieties that are ¦ 10 neutral or positively cha-ged.

In addition to the improvements in flux and molecular weight cutoff shown above the coated membranes of the invention perform differently than uncoated polysulfone UF
membranes with comparable molecular weight cutoff 1~ specifications. The data in Table 5 below show how a standa-d polysulfone membrane*(Udel) with a 2000 molecular weight cutoff exhibits lower flux and solute rejections than a coated UF membrane with an equivalent molecular weight cutoff. The rejection of the 2000 MW memb:ane appears to be determined by molecular size in both cases, while the rejection of the coated membrane appear dependent on a chemical interaction between the membrane surface and the feed solute, as may be derived from the following Table:
~ Trademark ~ ' ~ , I
~ . _ ~ij ~ ~

i~ , ~. ,' ` ~

`
-, -41- ~33232S
Table 5 Membrane Perfo~mance Data Comparison of a 2000 MW UF Membrane vs. a Coated UF Membrane (25 p.s.i.q.) DEXTRAN (2)5000 3ACITRACIN(3) S W~TER
FLUX FLUX REJECTION FLUX ~ ECTION
(GFD) (GFD) (~) (GFD) (%) 2000 MW (Udel) 30.1 24.8 76.5 7.6 50.3 Coated Membrane(l) 47. 7 37 O ~ 73.9 1.7 76.7 , ~','' ~
~ (1) 50,000 MW UF membrane ~Udel~ ~ ulfone) coated with sulfonated polyethersulfone (sulfonated ~ ,600).
(2) ~extr,~h 5000 is a linear polysaccharide with a molecular weight of 500~~
(3) 8acitracin is a globular protein with a lecular weight of 1400.
:
In application tests of the sulfonated polysulfone-coated membranes, the surface effect of coating a relatively hyd-ophobic polysulfone or similar subst_ate UF
membrane can alter the filtration characteristics dramat-ically. For example, uncoated commercial grade of poly-sulfone and polyamide UF membranes were severely fouled by a fermentation broth containing an alkaline protease and a defoamer. Surp-i ~ gly, a coated (sulfonated V~ct e~
polymer on a ~d ~ subst~ate) UF membrane of the invention showed a signlficantly higher flux and much less flux decay when treating the same alkaline protease fermenta-tion broth. Table 6 below compares the performance of coated membranes of the invention with standard indust-ial grade membranes on the concentration of rennen.
, ~ . , ',~
r '. ~
~ ''' ,S ~,'~
i~ ~
~i ~
i.i ''i'~'' ''';"' '' ' ' ' ' ' -42- 1~3232~ ~ ~

Table 6 -:
Evaluation of Sulfonated PolYethersulfone (Victrex 600) Coated UF ~ranes vs. Uncoated Membranes on Clarified Broth - Microblal Rennen (25 p.s.i.g.) PS~TE
I FLUX REJECTION
MEMBRANE TYPE ( GFD) UNITS
Coated Udel membrane (2,000 MW) 6 <100 Coated Udel membrane (10,000 MW) 6 <100 Uncoated S,000 MW Udel 9 100 Uncoated 10,000 MW Udel 12 200 Coated Udel 50,000 MW substrate12 300-600 Coated Udel 10,000 MW substrate12 1100 In summary, the application of a sulfonated polyaryl-ether sulfone coating to a UF membrane (substrate) has been found to provide the following advantages:

(a) Increased hydrophilicity - the -SO3H moieties in the ultrathin coating significantly inc-eases the hydrophilicy of the membrane surface, whereas : 20 unmodified polysulfone (and similar composition) UF
membranes are relatively hydrophobic; and, therefore, are more prone to fouling when exposed to oleophilic mate~ials.
:; ' (b) Charged Surface - the -SO3H moieties impart a strong negative charge to the membrane surface and :~ reduce normal mem~rane fouling tendencies since most colloidal materials are negatively charged and the SO3H groups will electrically repell them to keep : the membrane surface clean.

~30 . (c) Chemical reactivity - since the SO3H moiety is ~; extremely reactive this moiety can be easily ~::
~:
~: ~
:

~ . _ . ~

43 ~ 33232S

converted ,to other chemical moieties that will impart either neutral or cationic character ~o the membrane surface.

(d) Pore size control - the pore size distribution of the UF composite membrane can be regulated very closely by controlling either the key parametérs of the process, coating composition and curing cycle, or the porosity of the UF membrane substrate.

::

~ .
~ `

:
':

' :
.: ~ :

~,i,;,~,,, ,, ;-, , ,`, ~ ,,,, ,~"
7~
,~,, :,, -, -- : ~ - :
:, , -. : , - .
s ''' ~
r 4 .
:~' .
~' ` ~

Claims (16)

1. An improved oxidatively resistant membrane consisting essentially of an oxidatively resistant porous substrate to which is bonded a coating film of a sulfonated polyarylether polymer and which membrane is prepared by a process which comprises:
(a) forming a solution of a sulfonated polyarylether polymer in a potentiating solvent system containing at least 10%
formic acid, said solvent system being substantially polar, volatile enough to be removed by gentle heating, of low enough surface tension to wet the porous substrate, and not able to dissolve the porous substrate;
(b) uniformly applying said solution of the sulfonated polyarylether to at least one surface of the porous substrate;
and (c) removing said solvent from the solution to form a coating or film of the sulfonated polyarylether adherently attached to the porous substrate.

2. The membrane of claim 1 wherein the potentiating solvent system comprises:
(a) at least 20% to 100% formic acid;
(b) from 0% to 80% of a solution consisting essentially of (i) water and (ii) alkali metal salts.

3. The membrane of claim 2, wherein said potentiating solvent system contains minor amounts of alkali metal salts.

4. The membrane of claim 1 wherein the potentiating solvent system comprises at least 20% formic acid with the remainder of the solvent selected from the group consisting of an alcohol, alkylene diol or triol and an alkylene glycol alkyl ether.

5. The membrane of claim 1 wherein the potentiating solvent system comprises a lithium salt in an amount below about 2.0% by weight of the potentiating solvent system.

6. The membrane of claim 1 wherein the porous substrate is an anisotropic polymeric membrane comprising a polymer selected from the group consisting of polyarylether sulfones, polyarylether ketones, polyphenylene ethers, polyphenylene thioethers, polyimides, polyetherimides, polybenzimidazoles, polyesters, polyvinylidene fluoride, polychloroethers, polycarbonates, polystyrene, polyvinylchlorides, and polyacrylonitrile and various copolymers thereof.

7. The membrane of claim 1 wherein the porous substrate is an anisotropic polyarylether sulfone membrane.

8. The membrane of claim 1 fabricated for reverse osmosis applications wherein the thin coating or film of sulfonated polyarylether sulfone is capable of rejecting sodium chloride from a 3000 ppm aqueous solution thereof at a rate of at least 95% at 400 psig and yielding a flux of at least 15 GFD.

9. The membrane of claim 8 in the form of a spiral wound element or module.

10. The membrane of claim 1 fabricated for low pressure reverse osmosis or ultrafiltration applications wherein the porous substrate is an anisotropic polysulfone or polyethersulfone polymeric membrane.

11. The membrane of claim 10 in the form of a hollow fiber or tube module.

12. The membrane of claim 1 wherein the sulfonated polyarylether coating or film has an IEC of between about 0.2 and 2.0 milliequivalents/gm.

13. The membrane of claim 1 wherein the thin polymeric coating or film comprises a sulfonated polyarylether sulfone of the formula:
wherein Z is hydrogen, an alkali metal, or nitrogen containing species derived from ammonia or amines, the b/a ratio is from 0 to 20 and x is an average number such as to give an IEC for the sulfonated polymer of between 0.2 and 2.0 meq./gm.

14. The membrane of claim 1 wherein the thin polymeric coating or film comprises a sulfonated polyarylether sulfone of the formula:
wherein Z is hydrogen, an alkali metal or nitrogen containing species derived from ammonia or amines and x is an average number such as to give an IEC for the sulfonated polyarylether sulfone polymer of between 0.2 and 2.0 meq./gm.

15. The membrane of claim 1 wherein the thin polymeric coating or film comprises a sulfonated polyarylether ketone of the formula:
wherein Z is hydrogen, an alkali metal, or nitrogen containing species derived from ammonia or amines and x is an average number such as to give an IEC for the sulfonated polyarylether ketone polymer of between 0.2 and 2.0 meq./gm.

16. The membrane of claim 1, wherein said potentiating solvent system is capable of swelling the porous substrate.

(17) A process for making an improved sulfonated polyarylether coated or composite membrane on a porous polymeric membrane substrate which comprises (a) forming a solution of a sulfonated polyarylether polymer in a potentiating solvent system containing at least 10%
formic acid, which is substantially polar, volatile, of low enough surface tension to wet the porous substrate, and, though capable of swelling the porous substrate, is not able to dissolve it, (b) uniformly applying said solution of the sulfonated polyarylether to at least one surface of the porous substrate and (c) removing said solvent from the solution to form a coating or film of the sulfonated polyarylether adherently attached to the porous substrate.

(18) The process of claim 16 wherein the polymeric substrate is an anisotropic membrane comprising a polysulfone or a polyethersulfone and the sulfonated polyarylether film or coating has an IEC between 0.2 and 2.0 meq./gm.

(19) The process of claim 17 wherein the potentiating solvent system further comprises a minor amount of a lithium salt.

(20) A method for desalinating a saline water which comprises passing said saline water at an elevated pressure equal to or exceeding the osmotic pressure of said saline water over the oxidatively resistant membrane of claim 1.

(21) A method for ultrafiltering or microfiltering a fluid mixture which comprises passing said fluid mixture at moderately elevated pressure over the oxidatively resistant membrane of claim 1.